工程训练中心基于 OpenClaw 开源智能体开发平台进行本地化部署与深度适配， 核心技术与工具全部基于国产化方案：以 DeepSeek V4 + Kimi K3 大模型为底座， 融合学校"电、碳、辽"特色数据（电路故障案例、数控加工工艺参数、电力系统运行数据等）， 构建专属工训大模型。这不是简单的"API 调用"，而是 自主可控、数据不出域的本地化部署， 团队具备全栈驾驭 AI 技术内核的工程能力。

Built on the OpenClaw open-source agent platform with full local deployment and deep customization. Powered by DeepSeek V4 + Kimi K3 with domain-specific training on power, carbon, and Liaoning industrial data — circuit fault cases, CNC parameters, power system operations. This is not an API call. It is indigenous, data-sovereign, locally-deployed AI.

AI 智能体贯穿教学全流程：两大能力维度（知识问答与文档检索 + 动手实操与现场指导）、 七项核心功能（操作引导 / 故障排查 / 回路分析 / 虚实孪生 / 案例复现 / 实战推演 / 作业评阅）、 智能体能力层（个性化学习记忆 / 工具集成 / 主动学习引导 / 学习数据分析）—— 让产品 "越用越懂你"。

The AI agent spans the entire training pipeline: two capability dimensions (knowledge Q&A + hands-on guidance), seven core functions (guidance / fault diagnosis / circuit analysis / digital twin / case replay / scenario drill / assignment review), agent intelligence layer (personalized memory / tool integration / proactive guidance / learning analytics) — it knows you better over time.

未来规划：以"工训智脑"为起点，联合高校共建 医训智脑、建训智脑、矿训智脑， 联合企业共建 运维智脑、制造智脑， 构建覆盖 工科 · 医科 · 建筑 · 矿业 · 能源 · 制造 六大领域的 辽宁高校工程训练 AI 云平台。

Future roadmap: starting from "Engineering Training Brain", we will co-build Medical Training Brain, Construction Training Brain, Mining Training Brain with partner universities, and Operations Brain, Manufacturing Brain with industry partners — covering Engineering · Medical · Construction · Mining · Energy · Manufacturing on a unified Liaoning Higher Education Training AI Cloud Platform.